Can I use Html Agility Pack To Parse HTML Fragment?

If it's html then yes.

string str = "<b>Some code</b>";
// not sure if needed
string html = string.Format("<html><head></head><body>{0}</body></html>", str);
HtmlDocument doc = new HtmlDocument();
doc.LoadHtml(html);

// look xpath tutorials for how to select elements
// select 1st <b> element
HtmlNode bNode = doc.DocumentNode.SelectSingleNode("b[1]");
string boldText = bNode.InnerText;

I dont think this is really the best use of HtmlAgilityPack.

Normally I see people trying to parse large amounts of html using regular expressions and I point them towards HtmlAgilityPack but in this case I think it would be better to use a regex.

Roy Osherove has a blog post describing how you can strip out all the html from a snippet:

  • http://weblogs.asp.net/rosherove/archive/2003/05/13/6963.aspx

Even if you did get the correct xpath with Mika Kolari's sample this would only work for a snippet with a <b> tag in it and would break if the code changed.